5401
J . Am. Chem. SOC.1980, 102, 5407-5409 Table 1. Summary o f Spectral Data for Compounds 9, 12,a n d 13 compd
9
12
13
mass spectra, m / e (re1 int)
UVa Amax, nm
(€1
I R , ~cm-’ (int)
218 (18, M+), 189 (54), 226 (7300),239 (5940), 2190 (w), 2140 (s), 161 (86),133 (100) 251 (6900),268 sh (8420), 2045 (s), 1615 (m), 300 (22000),318 (209001, 730 (s) 368 (175), 398 (53)
H NMR,‘ 6 0.3-1.5 (15 H,m, Si-CH,CH,), 2.04 (6 H,S, CH,), 5.27 (1 H,br s, CH)
,c N M R , s~ 166.04 (C-41,150.35 (c-7), 122.01 (C-5),103.66(C-6), 94.57(C-3),81.81 (C-8), 24.59,24.45 (C-1,C-2), 7.26 (C-lo),4.33 (‘2-9)
0.97 (1 H,s, 186 (51, Mt), 171 (18), 215.5 (42300),219 (41 3001, 3080 (w), 2215 (s), 2130 (w), 1603 (m), cyclopropyl), 156 (29),141 (52), 243 (5370),256 (8470), 1.18 (12 H,s, CH,), 83 (100) 270.5 (1 1 IOO), 287 (8290) 898 (s) 1.92(3 H,m,CH,), 5.40 (2 H,m, C=CH,) 95.80 (C-3),88.47 (C-61, 368 (25,M+), 353 (33), 244 (2210),257.5 (21901, 3080 (w), 2215 (s), 0.83 (4H,m, cyclopropyl), 75.30 (C-4),69.86 (‘2-51, 339 (17),325 (38), 272.5 (1540) 2140 (w) 1.26 (12 H,S, CH,), 42.15 (C-2),29.46 (c-71, 31 1 (24),297 (59), 23.41 (C-l),20.98, 20.30, 1.10-1.55 (8 H,m, 283 (60),269 (811, cyclohexane), 19.69 (C-8,C-9,C-10) 255 (100) 1.55-2.00 (8 H,m, cyclohexane)
9 a n d 12 i n hexane, I3 in EtOH.
9 a n d 12 neat, 13 10%in CCI,.
spectral properties are fully consistent with the proposed structures and the available literature analogies. In particular, all three compounds gave good molecular ions and easily recognized fragment ions in the mass spectrometer. The intense UV absorptions at 300 and 318 nm and the strong infrared bands at 2045 and 2140 cm-I are highly characteristic of hexapentaenes’ and hence in accord with the proposed cumulene 9, as is its proton N M R . The most definitive spectral feature of cumulene 9 is its I’C N M R and the low-field absorptions between 81 and 166 ppm due to the six unsaturated carbons characteristic of such sp-hybridized cumulene carbons.’~~ The identity of 12 is unambiguously established by its ‘H NMR and confirmed by its UV and IR absorption typical of such d i y n e ~ . ~ The structure of dimer 13 is established by its molecular ion at m l e 368 and its I3C N M R with four distinct acetylenic carbons in the 69-96-ppm range, as well as by its UV and IR absorption, all in excellent agreement with known cyclyne analogues 141° and 15.” Unfortunately, neither the spectral data nor the mode of
/tR R
14, R = H 15, R = CH, formation allow a distinction between the two possible cyclodimers, 13a or 13b. However, the fact that only ten signals were observed in the I3C N M R and that only a single sharp spot is seen on TLC strongly suggests that only one of the two possible dimers of 13 is actually formed. In summary, we have established that highly unsaturated extended carbenes (6) may be readily generated and trapped, albeit in only modest yields. Insertion of 5 into Et3SiH gives previously unknownS novel silicon-functionalizedcumulene 9 whereas addition to olefins results in rearranged ene-diyne 12 or an unusual dimer, namely, cyclyne 13. The full scope of alkatetraenylidenecarbene generation as well as the nature and chemistry of these reactive intermediates are under active investigation. (8) Van Dongen, J. P. C. M.; de Bie, M. J. A,; Steur, R. Tetrahedron Lett. 1973, 137 1-1 374.
(9) Patai, S.,Ed. “The Chemistry of the Carbon-Carbon Triple Bond”, Wiley-Interscience: London, 1978; Parts 1 and 2. (10) Wolovsky, R.; Sondheimer, F. J . Am. Chem. SOC. 1965, 87, 5720-5727. (11) Scott, L. T.; DeCicco, G. J. Tetrahedron Lett. 1976, 2663-2666. Santiago, C.; Houk, K. N.; DeCicco, G. J.; Scott, L. T. J . Am. Chem. SOC. 191a,100,692-696. (12) Fulbright-Hays postdoctoral fellow from the University of Zagreb, Zagreb, Yugoslavia.
0002-7863/80/1502-5407$01 .OO/O
CCI,, internal Me,Si.
CDCI,, proton decoupled.
Acknowledgments. This research was supported by the National Science Foundation (CHE78-03596). Peter J. Stang,* MIaden Ladika Chemistry Department The University of Utah Salt Lake City, Utah 841 12 Received March 27. 1980 I
ESR and IR Evidence for Intermediates in the Di- a-methane Photorearrangemen t of a Naphthobarrelene-Like Compound after Low-Temperature UV Irradiation Sir: We previously reported on the photorearrangement of the bicyclo[3.2.2]nonanaphthalene 1 to give three isomeric semibullvalene-like products 6a-c.1 A similar rearrangement also occurs, in a reversible fashion, in a thermal ground state process.’ It was proposed that the di-a-methane photorearrangement 1 --* 6a-c proceeds along reaction paths which evolve from two primary photochemical processes, namely naphthyl-vinyl (-2) and vinyl-vinyl bonding (-4). Evidence for competition between several paths and the involvement of a transient ( T 1.5 X s at 280-350 K) in addition to excited triplet 1 was presented.’ We now present direct proof of the existence of two discrete intermediates in the transformation 1 + hu 6a-c obtained by low-temperature ESR and IR spectroscopy. Irradiations were carried out in 2-methyltetrahydrofuran (MTHF)3 matrixes with 313- and >340-nm light at 77 K.4 Continuous irradiation at either excitation wavelength gave two ESR spectra. A transient spectrum showed a strong Am = 2 transition at 1764 G and a Am = 1 portion ( X Y Z pattern) with D ’ = 800 G (D= 0.075 cm-I) and E ’ = 66 G ( E = 0.0065 cm-I). In addition, a partially line-resolved signal dominated at 3244 G (=Ho). Unlike the other signals, the latter persisted when irradiation was discontinued, without a decrease in intensity over a period of several hours. The persistent central signal could be resolved into the spectrum given in Figure 1. The zero-field parameters are D’ = 58 G ( D = 0.0054 cm-l)
-
-
(1). Demuth, M.; Bender, C. 0.;Braslavsky, S. E.; Gorner, H.; Burger, U.; Amrein, W.; Schaffner, K. H e h . Chim. Acta 1979, 62, 847-851. (2) Demuth, M.; Burger, U.;Mueller, H. W.; Schaffner, K. J . Am. Chem. SOC.1979, 101, 6763-6765. (3) In the sequence hexafluorobenzene, ether/isopentane/ethanol (5:5:2), and MTHF, the ESR signal resolution increased markedly. (4) Light source: SP-1000 W high-pressure Hg lamp (Philips Richtstrahler combined with liquid filter solutions [313 nm, 0.567 g of K2Cr04and 2.056 g of Na2C03in 1 L of distilled H20, >340 nm, 750 g of NaBr and 8 g of Pb(NO& in 1 L of H20]. Varian E9 ESR spectrometer: sensitivity 5 X 1Olo of AH spins, spectra of samples [(1.0-6.6) X lo-* M I ] placed in a liquid nitrogen cooled Dewar were recorded at 9.05-GHz microwave frequency. Perkin-Elmer 580 IR spectrometer: IO-] M 1, 25-pm path length.
0 1980 American Chemical Society
5408 J . Am. Chem. SOC., Vol. 102, No. 16, 1980
Communications to the Editor
Scheme I. Reaction Paths Proposed for the Photorearrangement 1 +6a-c’
9
@ \
A ESR signal of biradical at 77K
25
0 Product formation after warming
to room temperature 4 w
1
10
!
l
l
1
20 30 !rradiation time. mip
1
LO
~
1
50
Figure 2. Comparison of long-lived ESR signal intensity and formation of product 6 as a function of irradiation time. (A) Excitation at 313 nm; (B) Excitation a t >340 nm. (0)Amplitude of ESR Y line; (A)product 6 formation after the ESR sample was warmed to room temperature. Maxima of ESR intensity and product formation taken as 100%. Note that the diagram is not normalized for light intensities of the two exci-
\
2
l
4
tation wavelengths.
,;“
before irradiation
77K
I
3 n
A
I
n
after irradiation 77 K
starting materia: (1 1
92 K intermediate(s1
hv
hv
6a
6b
6c
I
182K
w1700
165C
product ( 6 )
cm-’
Figure 3. IR spectra recorded at 77 K before irradiation of 1 and after irradiation of 1 for 30 min during warming up from 77 to 242 K. Solvent
Figure 1. Long-lived component of the ESR spectrum, recorded in the dark after irradiation of 1 at >340 nm in 2-methyltetrahydrofuran matrix at 77 K. The shaded portion corresponds to radical byproducts.
and E‘= 4 G ( E = 0.0004 cm-I). A Am = 2 transition was not observed even under high gain and modulation adjustment. The parameters of the short-lived component of the ESR spectrum are characteristic of an electronically excited and randomly oriented aromatic triplet state,5 which can be attributed to the naphthalene of 1 The persistent spectrum (Figure 1) has all the characteristics expected of a randomly oriented ground-state triplet biradi~al.’*~ The lack of a half-field signal (Am = 2) is in accord with the small zero-field parameter, characteristic of a weak interaction between the radical centers. ( 5 ) Wasserman, E.; Snyder, L. C.; Yager, W. A. J . Chem. Phys. 1964,4/, 1763-1772. (6) Note that 1 has a naphthalene-like phosphorescence at 77 K: Amrein, W.; Schaffner, K. H e h . Chim. Acta 1975, 58, 397-415. For the ESR parameters of triplet naphthalene, see ref 7. (7) Wan, J. K. S. Adu. Phorochem. 1974, 9, 1-145, and references therein. (8) Wertz, J. E.; Bolton, J. R. “Electron Spin Resonance, Elementary Theory and Practical Applications”, McGraw-Hill: New York, 1972; Chapter 10.
2methyltetrahydrofuran, X(excitation) >340 nm. Note that the band shapes of 1 (dark portions) and 6 (shaded) at 242 K are the same as those of the respective reference samples at this temperature. Only one frozen-out rotamer of 1 appears at 77-182 K after irradiation,” and the buildup of 6 begins to show at 182 K.
-
W e assign the biradical spectrum to an intermediate (or intermediates; see below) in the rearrangement 1 6a-c. No ESR signal was found upon irradiation of 6 a t either wavelength in MTHF at 77 K. The Y-line amplitude’ of the long-lived signal reached maximum intensity after -35-min irradiation at 313 nm and after -15-min irradiation at >340 nm. Since the signal intensity did not decrease in the dark after irradiation, a comparison with product formation is meaningful. The amount of 6 formed was measured by GLC after the ESR samples were warmed to room temperature. Normalizing maximum product formation and ESR signal saturation to 100% enabled the relative increases in product formation and ESR signal intensity with time to be compared (Figure 2). The close fit of the two curves for both excitation wavelengths is compatible with the role of the biradical as a precursor to 6 . The possibility that the biradical is an intermediate leading to another product (rather than to 6) can be eliminated. For a line width of 12 G (Y line), a minimum M is required for detection steady-state concentration of ca. (9) The Y line was chosen in view of its large amplitude, symmetric shape, and independence from the zero-field monoradical absorption.
J . Am. Chem. SOC.1980, 102, 5409-5410 of any ESR signal with our e q ~ i p m e n t .Yet, ~ no products other than 1 and 6 were found by GLC, under conditions where a product concentration 1 5 X lo-’ M could have been detected.” An IR investigation showed the consecutive appearance of new frequencies (I, 1634 and 1648-1655 cm-I; 11, 1660 cm-I) upon irradiation for 30 min at 77 K and warming in the dark to 92 K, before the formation of product 6 (vco = 1665 cm-l) (Figure 3).” A three-step sequence is thus established which can be accommodated by the proposed reaction scheme. Both the long-lived ESR component and the transient IR bands I are observable only below 92 K, Le., they may well arise from the primary metastable photoproduct(s), either 2 alone or a mixture of 2 and 412 (note that the ratio (6a + b)/6c is -5:l at 77 K’). If so, then the IR band I1 could be due to 3 (and possibly 5). The IR frequencies I and I1 are suggested to be those of the a-keto (oxallyl) radical groups, although we are not aware of appropriate reference data. The observed D value of 0.0054 cm-* is considerably smaller than any of the D values reported previously for triplet biradicals. Our assignment to the extensively delocalized biradical 2 appears acceptable, however, in light of the spread of D from 0.01-0.03 cm-l for partly delocalized biradicals14 to 0.08 cm-’ for the localized 1,3-cyclopenta diyl.I5 The transition from intermediate 2 to 3 when the temperature is raised to 92 K should be accompanied by the development of a new ESR spectrum which then should persist until 180 K. However, no resolved signal was observed when measurements were carried out at 92 K with a flow cavity. Although it cannot be ruled out that this result was merely due to inhomogeneity caused by the nitrogen flow, the assignment of structure 3 to the species responsible to the IR band I1 must therefore remain tentative. Still, it appears more plausible than the alternative that the IR change between 77 and 92 K is associated with a rotation of the benzoyl group in 2 from out of conjugation into conjugation with the radical center. If this were the case, one might expect a shift to lower rather than higher IR frequency (as is observed in Figure 3). Furthermore, the benzoyl rotamers of 1 are frozen out at ca. 200 K, and oxallyl rotamers (as in 2-5) are likely to have a higher barrier. In conclusion, we note that, for the first time in di-a-methane photochemistry, direct evidence for the intervention of a biradical intermediate (or a category of biradical intermediates) and strong support for a second intermediate have been obtained. This result confirms the postulated16 stepwise mechanism for the photorearrangement of barrelenes. Furthermore, it corroborated the conclusions previously drawnI3 from the rearrangement of independently generated biradical intermediates of the type now observed directly.
-
5409
Acknowledgments. We thank W. Riemer for able technical assistance, and we greatly appreciate stimulating criticism by Professor P. J. Wagner. Martin Demuth, Dieter Lemmer, Kurt Schaffner* Institut fur Strahlenchemie im Max-Planck- Institut fur Kohlenforschung 0-4330 Miilheim a. d. Ruhr, West Germany Received December 28, 1979
Effect of Through-Bond Interaction on Terminal Methylene Rotation in the Tetramethylene Diradical Sir: How the ease of terminal methylene rotation in tetramethylene (butane-1 ,4-diyl) depends upon the conformation about the central C-C bond remains a major unanswered question in the study of this diradical.] The pioneering investigation by Hoffmann and co-workers2 of the potential surface for the singlet diradical revealed small barriers to terminal methylene rotation in the gauche conformation about the central C-C bond but large barriers in the trans conformation. In the latter conformation, the strong perference found for the 90,90 geometry at the terminal methylene groups (Figure 1) was traced to the stabilizing interaction between the formally nonbonding atomic orbitals (AO’s) on these carbons and the orbitals of the central C-C bond. As a result of this “through-bond coupling” between the AO’s on the terminal carbons in the 90,90 geometry of the trans diradical, Hoffmann’s EH calculations found the 90,90 geometry to be favored over 0,90 and 0,O by, respectively, 12.7 and 13.4 kcal/mol. From INDO calculations, Stephenson and Gibson3 obtained values of 45.0 and 46.7 kcal/mol for these energy differences. Both of the above calculations are flawed by the use of oneconfiguration wave functions to represent singlet diradicals. In order to reduce the contribution of high-energy ionic terms to the wave function for a homosymmetric singlet diradical, a twoconfiguration wave function is r e q ~ i r e d . ~For instance, if the closely spaced 7a, and 6b, molecular orbitals (MO’s) at the 90,90 tetramethylene geometry are denoted as S and A, corresponding to the symmetry behavior of each with respect to the twofold axis, the wave function for the lowest singlet state takes the form of eq 1.
‘A, = c I ( (IO) Glass capillary GLC column; OV 101, 70 ft. Components >0.05% were detectable, Le., 5 X IO” M at a total concentration of lo-* M. A further factor of ten derives from the minimum ESR signal detectability for a 10% amplitude in the plot of Figure 2, safely assuming a maximum biradical concentration of lo-’ M for 100% signal amplitude. (1 1) The IR experiment also reveals the striking fact that of the two rotamers of 1 which are frozen out at 77 K the major component (vCo = 1633 cm-I) photoreacts preferentially while the minor component (vco = 1644 cm-I) remains largely unchanged (accounting for ca. 30% recovered 1 after 2-h irradiation at 77 K). We note that the residual IR band at around 1647 cm-’ (242 K) exactly corresponds to the carbonyl band of 1 under these conditions. This remarkable influence of excited-state geometry, as reflected by the ground-state conformation, on the photoreactivity at 77 K is currently under further investigation. At room temperature, a similar effect is not observed for product formation = 1.O1). (12) The intermediate 4 is structurally similar to a triplet biradical which has been generated by diazo decomposition and shown to preferentially afford the di-*-methane product of 2,3-naphthobarrelene.l3 (13) Zimmerman, H. E.; Boettcher, R. J.; Buehler, N. E.; Keck, G. E.; Steinmetz, M. G. J. A m . Chem. SOC.1976, 98, 7680-7689. (14) Dowd, P. J . A m . Chem. SOC.1966, 88, 2587-2589; 1970, 92 1066-1068. Berson, J. A.; Bushby, R. J.; McBride, J. M.; Tremelling, M. Ibid. 1971, 93, 1544-1546. Watson, C. R., Jr.; Pagni, R. M.; Dodd, J. R.; Bloor, J. E. Ibid. 1976, 98, 2551-2562. Roth, W. R.; Erker, G. Angew. Chem’. 1973,85, 510-511; Angew. Chem., Int. Ed. Engl. 1973, 12, 503-504. (15) Buchwalter, S. L.; Closs, G. L. J . A m . Chem. SOC.1979, 101, 4688-4694. (16) Zimmerman, H. E.; Binkley, R. W.; Givens, R. S.;Grunewald, G. L.; Sherwin, M. A. J . Am. Chem. SOC.1969, 91, 3316-3323.
0002-7863/80/ 1502-5409$01 .OO/O
. . . A’)
-~
2 1.
. . S2)
In eq 1, c1 > c2 since A lies below S in energy. The origin of this energy difference is that the interactions in A between the AO’s on the terminal carbons and the orbitals of the central C-C linkage are bonding while in S these interactions are antibonding2 Consequently, the presence of 1 . . . S2)in the two-configuration IA, wave function might be expected to reduce the net bonding between the AO’s on the terminal carbons and the orbitals of the central C-C bond, thus lowering the energy difference between 90,90 and both 0,90 and 0,O. The 90,90 geometry of trans-tetramethylene is a true intermediate on the STO-3G CI surface calculated by Segal for the singlet d i r a d i ~ a l . However, ~ Segal did not compute the energies of the 0,90 and 0,O geometries. In order to determine whether ~
( I ) For recent reviews of tetramethylene chemistry, see: Borden, W. T. React. Intermed. 1980, in press. Berson, J. A. In “Rearrangements in Ground
and Excited States”; de Mayo, P., Ed.; Academic Press: New York, 1980; p 311. (2) Hoffmann, R.; Swaminathan, S.; Odell, B. G.; Gleiter, R. J . A m . Chem. SOC.1970, 92, 7091. (3) Stephenson, L. M.; Gibson, T. A. J . Am. Chem. SOC.1972, 94,4599. (4) See, for instance: Salem, L.; Rowland, C. Angew. Chem., In?. Ed. Engl. 1972, 11, 92. (5) Segal, G. A. J . A m . Chem. SOC.1974, 96, 7892.
0 1980 American Chemical Society
